DES was announced as a standard in 1976. AES competition started in 1997 and Rijndael was selected as standard in 2000. What are major block ciphers and block cipher designs made/proposed from 1976 to 1997?


IDEA and CAST-128 (aka CAST5) first come to my mind as widely deployed/used ciphers. Blowfish and TEA are next. The first two (and to some degree the third) got support in PGP early on.

Update per comment: RC2, RC5, and GOST 28147-89 (in the Eastern block) also have been widely used.

All 7 are 64-bit block ciphers; that's typical of the era, and made these block ciphers a painless substitute for DES and 3DES (the dominant ciphers). They stood the test of time quite well, with no practical attack better than brute force by more than very few bits (except for GHOST, but when used with its full 256-bit key it remains safe enough). While TEA is involved in a notorious incident where a game console misused it for hashing, that's hardly a problem attributable to TEA itself.

Note: I'd have included RC4 if is was a block cipher; that used to be widespread in quite a few commercial applications.

| improve this answer | |
  • 2
    $\begingroup$ PES (successor of IDEA) and FEAL also come to mind. Square (1997) was definitely important (origin of integral attacks). Skipjack is probably also relevant (but was classified until 1998). Same goes for GOST. Of course, the length of the list depends on what we will consider as "major". e.g. many designs succeeded Rijndael (including Square, 3-Way and BaseKing). $\endgroup$ – Aleph Feb 12 '18 at 19:44
  • 1
    $\begingroup$ @Aleph: Wikipedia says PES is an early version of IDEA. I've heard more about attacks than uses regarding FEAL. $\endgroup$ – fgrieu Feb 13 '18 at 8:10
  • 2
    $\begingroup$ RC4 may not be applicable, but RC2 and RC5 definitely are. RC2 was designed specifically to be used in Lotus Notes. On the Soviet side, GOST 28147-89 was pretty major. Merkle's Khufu and Khafre are noteworthy as early software-friendly designs, but unclear what their actual impact was. $\endgroup$ – Samuel Neves Feb 13 '18 at 8:39
  • $\begingroup$ @fgrieu Should have been "predecessor" rather than "successor" (same for Square, 3-Way and BaseKing). I'd argue that being used isn't necessarily the only argument for importance. $\endgroup$ – Aleph Feb 13 '18 at 8:43
  • 2
    $\begingroup$ I'd include 3DES in the list. $\endgroup$ – CodesInChaos Feb 13 '18 at 9:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.